Spaces:
Sleeping
Sleeping
| import streamlit as st | |
| original_title = '<p style="font-family:Trebuchet MS; color:#000000; font-size: 25px; font-weight:bold">ASCARIS: Positional Feature Annotation and Protein Structure-Based Representation of Single Amino Acid Variations</p>' | |
| st.markdown(original_title, unsafe_allow_html=True) | |
| text = '<p style="font-family:Trebuchet MS; font-size: 20px; font-weight:bold">Developers: Fatma Cankara & Tunca Dogan</p>' | |
| st.markdown(f'<p class="title-text">{text}</p>', unsafe_allow_html=True) | |
| st.markdown(""" | |
| <style> | |
| .main-text | |
| { | |
| font-family:Trebuchet MS; font-size:14px;text-align: justify;font-weight:bold | |
| } | |
| </style>""", unsafe_allow_html=True) | |
| text = 'ASCARIS (Annotation and StruCture-bAsed RepresentatIon of Single amino acid variations) is a tool for the featurization (i.e., quantitative representation) of single amino acid variations (SAVs), which could be used for a variety of purposes, such as predicting their functional effects or building multi-omics-based integrative models. ASCARIS utilizes the correspondence between the location of the SAV on the sequence and 30 different types of positional feature annotations (e.g., active/lipidation/glycosylation sites; calcium/metal/DNA binding, inter/transmembrane regions, etc.) from UniProt, along with structural features and the change in physicochemical properties, using models from PDB and AlphaFold-DB. It constructs a 74-dimensional feature set (including meta-data) to represent a given SAV.' | |
| st.markdown(f'<p class="title-text">{text}</p>', unsafe_allow_html=True) | |
| text = 'Please refer to our pre-print article for more information on the construction of feature vectors, statistical analysis of features, and machine learning models trained on ASCARIS representations to predict the effect of SAVs:' | |
| st.markdown(f'<p class="title-text">{text}</p>', unsafe_allow_html=True) | |
| st.markdown(f'<p class="title-text"></p><a href="https://www.sciencedirect.com/science/article/pii/S200103702300329X">Cankara, F., & Doğan, T. (2023). ASCARIS: Positional feature annotation and protein structure-based representation of single amino acid variations. Computational and Structural Biotechnology Journal, 21, 4743-4758.</a>', unsafe_allow_html=True) | |
| st.write('') | |
| st.write('') | |
| st.image('visuals/concept_figure.png') | |
| text = '<p style="font-family:Trebuchet MS; font-size: 20px; font-weight:bold">Description of the Dimensions of ASCARIS SAV Representations</p>' | |
| st.markdown(f'<p class="title-text">{text}</p>', unsafe_allow_html=True) | |
| st.markdown(""" | |
| <style> | |
| .main-text | |
| { | |
| font-family:Trebuchet MS; font-size:14px;text-align: justify;font-weight:bold | |
| } | |
| </style>""", unsafe_allow_html=True) | |
| text = 'In ASCARIS representations, dimensions 1-5 correspond to datapoint identifier, 6-9 correspond to physicochemical property values, 10-12 correspond to domain-related information, 13-14 correspond to information regarding variation position on the protein (both the sasa value and the categorization), 15-44 correspond to binary correspondence between the variation and different types of positional annotations (1 dimension for each annotation type, for a total of 30 types), 45-74 correspond to spatial (Euclidian) distances between the variation and different types of positional annotations (1 dimension for each annotation type, for a total of 30 types).</p>' | |
| st.markdown(f'<p class="title-text">{text}</p>', unsafe_allow_html=True) | |
| st.markdown("""| Order of dimension | Column name in the output file | Description | Source | | |
| | ------------- | ------------- | ------------- | ------------- | | |
| | 1 | prot_uniprotAcc | UniProt accession | Metadata obtained from UniProtKB/Swiss-Prot | | |
| | 2 | wt_residue | Wild type residue | Data obtained from UniProtKB/Swiss-Prot (humsavar), ClinVar, PMD | | |
| | 3 | mut_residue | Mutated residue | Data obtained from UniProtKB/Swiss-Prot (humsavar), ClinVar, PMD | | |
| | 4 | position | Variation position | Data obtained from UniProtKB/Swiss-Prot (humsavar), ClinVar, PMD | | |
| | 5 | meta_merged | Datapoint identifier (UniProt accession-WT Residue-VariationPosition-Mutated Residue) | - | | |
| | 6 | composition | Change in composition values upon the occurrence of variation. Composition is defined as the atomic weight ratio of hetero (non-carbon) elements in end groups or rings to carbons in the side chain. | Literature | | |
| | 7 | polarity | Change in polarity values upon variation. | Literature | | |
| | 8 | volume | Change in volume values upon variation. | Literature | | |
| | 9 | granthamScore | Change in Grantham scores (the combination of composition, polarity and volume) values upon variation. | Literature | | |
| | 10 | domains_all | InterPro Domain IDs of all domains found in the dataset | Data obtained from InterPro | | |
| | 11 | domains_sig | InterPro Domain IDs of significant domains in the dataset. Domains that are not found to be significant in Fisher's Exact Test are labelled as "NULL". | Data obtained from InterPro | | |
| | 12 | domains_3Ddist | Shortest Euclidian distance between the domain and the variation site. | A newly engineered feature (data obtained from PDB/AlphaFold and InterPro) | | |
| | 13 | sasa | Solvent accessible surface area values. | FreeSASA | | |
| | 14 | location_3state | Caterozied location of the variation in the structure: surface, core or interface. | FreeSASA, InteractomeInsider | | |
| | 15-44 |disulfide_bin, intMet_bin,intramembrane_bin, naturalVariant_bin, dnaBinding_bin, activeSite_bin, nucleotideBinding_bin, lipidation_bin, site_bin, transmembrane_bin, crosslink_bin, mutagenesis_bin, strand_bin, helix_bin, turn_bin, metalBinding_bin, repeat_bin, caBinding_bin, topologicalDomain_bin, bindingSite_bin, region_bin, signalPeptide_bin, modifiedResidue_bin, zincFinger_bin, motif_bin, coiledCoil_bin, peptide_bin, transitPeptide_bin, glycosylation_bin, propeptide_bin | Positional sequence annotations, binary correspondence-based (30 different types of annotations, each one on a different dimension). Categories: 0: annotatation does not exist on the protein, 1: annotation is presented, but the variation is not on the annotated site, 2: variation is on the annotated site. | Newly engineered features (data obtained from UniProtKB) | | |
| | 45-74 |disulfide_dist, intMet_dist, intramembrane_dist, naturalVariant_dist, dnaBinding_dist, activeSite_dist, nucleotideBinding_dist, lipidation_dist, site_dist, transmembrane_dist, crosslink_dist, mutagenesis_dist, strand_dist, helix_dist, turn_dist, metalBinding_dist, repeat_dist, caBinding_dist, topologicalDomain_dist, bindingSite_dist, region_dist, signalPeptide_dist, modifiedResidue_dist, zincFinger_dist, motif_dist, coiledCoil_dist, peptide_dist, transitPeptide_dist, glycosylation_dist, propeptide_dist | Positional sequence annotations, distance-based (the spatial distance between the annotated residue and the mutated residue, in the protein structure, for 30 different types of annotations, each one on a different dimension), in terms of Angstroms. | Newly engineered features (data obtained from PDB/AlphaFold and UniProtKB) | | |
| """) | |